RESERVOIR DESCRIPTION
This article is based on Crain's Petrophysical Pocket Pal" by E. R. (Ross) Crain, P.Eng. Copyright 2018 by the author. Copying or distribution in any form without express written permission is prohibited.

Reservoir description, sometimes called reservoir characterization or reservoir modeling, attempts to create a static and a dynamic three-dimensional description of an oil or gas reservoir, based on the one- and two-dimensional data from well bores and seismic surveys. A good reservoir description costs money and requires an integrated, multi-discipline approach.

The direct aims of the static reservoir description are:
  1. Extrapolate core data to uncored wells
  2. Define quantity and distribution of porosity, saturation, and permeability in each well
  3. Interpolate rock property data between wells
  4. Identify flow units from porosity vs permeability populations
  5. Build a knowledge base that evolves with the reservoir development

The direct aims of the dynamic reservoir description are:
  1. Test the static model for accuracy by matching production history
  2. Predict future performance under various operational scenarios
  3. Optimize production for maximum long-term economic return

A large fraction of the data for the static model comes from the petrophysical analysis, along with the core and petrographic data used to calibrate the log analysis.

The beginnings of a reservoir description that will continue to evolve over time. During an earlier
iteration, the sand channel in the middle of the carbonate reservoir had not been discovered, so
the earlier model had to be completely revised.

Reservoir Description means many things to many people. Thousands of consultants, contractors, and oil company department heads use the term with subtle or serious differences in meaning.

The petrographer thinks reservoir description means defining the pore geometry, pore size, and pore throat radius distributions, and reservoir mineralogy, using thin section analysis, X-ray diffraction (XRD), and scanning electron microscopes (SEM). The modern term used is “porosity imaging”. The source material is drill cuttings or samples from cores. Their work is usually at the sub-millimeter level.

Core analysts think reservoir description is the measurement of porosity, permeability, grain density, capillary pressure, relative permeability, and electrical properties of the rocks, as well as facies description from observation of the depositional environment seen in slabbed core and core photographs. This work is at the centimeter level.

Petrophysicists see reservoir description as the evaluation of well logs to obtain reservoir rock and fluid properties, such as shale volume, porosity, water saturation, permeability, and lithology on a foot by foot basis, as well as sums and averages over specific reservoir units. This work is usually calibrated by core analysis and petrographic data where it is available. When well log data is combined with other geoscience data to form a coherent picture, it is called Integrated Petrophysics. Some work, such as analysis of formation microscanner images, may be at the centimeter level, but most is done at the tool resolution level – usually 0.3 to 1 meter.

Geophysicists see reservoir description as the creation and mapping of seismic attributes or inverted seismic data to illuminate variations in reservoir properties between well control. This work is at the multi-meter level vertically and horizontally, but provides finer spatial resolution than logs and cores, which are dictated by well spacing. Attributes are usually calibrated to petrophysical log analysis results.

Geologists perceive reservoir description as the interpretation and mapping of petrographic results, core analysis, petrophysical rock properties, and seismic attributes into stratigraphic sequences and/or flow units. In the simplest cases, the mapping is based solely on correlation of raw log curves. In more elaborate studies, all of the measured and computed data will be mapped. Dipmeter and pressure transient analysis results may be introduced to assist in correlation or definition of reservoir boundaries or fault planes. These maps are also at the multi-meter level vertically and either the seismic shot point spacing or well spacing areally. The geological model obtained is the static reservoir description which, of course, can be monitored and varied over time by acquisition of new petrology, core data, or petrophysical rock properties from new wells or logs run through casing. Time-lapse (4-D) seismic or through casing logs may also be used to monitor hydrocarbon contact changes.

Different geostatistical realizations of well log data suggest  different hydrocarbon recoveries.

Reservoir engineers view reservoir description as the analysis and interpretation of pressure transient data from drill stem tests or production tests. Flow capacity (permeability times thickness) and distance to reservoir boundaries (if they are close enough to be sensed by the pressure response) are the usual results obtained. In addition, pressure changes with production versus time provide grist for the material balance calculation mill. Production history (oil, gas, and water volumes versus time) coupled with decline curve analysis leads to predictions of ultimate hydrocarbon recovery. This work is also at the multi-meter level vertically and limited to the tested or produced wells only. Reservoir engineers are concerned with this dynamic reservoir description, as well as the static description for calculating reservoir volume and recoverable reserves.

Simulation engineers visualize reservoir description as a totally dynamic description of reservoir performance. They do use the static reservoir description (the geological model) as the basic foundation for the reservoir simulation. The bricks and mortar added to this foundation are the pressure and production data, and fluid properties, from the reservoir engineer. The object of the simulation is first to match production history, then to predict future behaviour of the reservoir. A good history match usually means that the geological model is reasonably accurate. The critical test is to compare the performance prediction with actual performance after a few years has elapsed. The reservoir simulation grid is usually a few to many meters vertically and many meters areally. Since more than one geological model could provide an adequate history match, calibration can only come with the passage of time and the arrival of new well data. Various production scenarios may be run in order to optimize reservoir recovery and economics.

Integrating the various geoscience disciplines is central to achieving a reasonable dynamic reservoir description.

Production and facilities engineers see the dynamic reservoir model as a template for design and economic evaluation of production, gathering, treating, and pipeline  equipment required to handle the predicted reservoir performance. The timing of compressor installations, water disposal wells, and conversion of wells to injection are paramount considerations. Both undersized and oversized facilities reduce the economic return from hydrocarbon production.

Drilling and production engineers use the various scenarios to plan in-fill drilling and re-completion operations on the wells in the reservoir.

Economic engineers use the dynamic and static models to predict cash flow, financing requirements, and investment decisions. Management and shareholders use the models to assist in making decisions as well.

The truth is that all these definitions are true, but the sum of the parts is even more true. Multi-disciplines generate the most complete reservoir description. Teamwork is everthing! Few people are talented enough to do it all without help from other specialists.

“Integrated” reservoir models are based on the multi-discipline approach to the model. Integrated reservoir description combines geology, geophysics, petrophysics, reservoir engineering, production engineering, and numerical simulation. All of the disciplines listed above talk to each other and iterate their interpretations until all the data fits a common definition of the reservoir.

For example, a history match may fail because there is not enough reservoir volume to account for production and pressure trends to date. Reservoir volume can be increased by changing the parameters and cutoffs that go into petrophysical calculations, or the geological mapping can be more generous in contouring the data. One of the biggest problems is up-scaling from the sub-millimeter pore distribution through all the steps to the multi-meter grid cells of a reservoir simulation model. An open mind and a willing heart might be needed to overcome a lot of problems!

FLOW UNITS
Flow units are the basic breakdown of a reservoir rock into layers. Each layer, or flow unit,  can be treated as a homogenous rock with uniform properties. A series of layers make up a zone, and a series of zones make up a reservoir.

A flow unit is a rock body with a distinct pore system or porosity-permeability distribution. Each flow unit responds differently to fluid flow and production. Defining flow units in sandstone reservoirs is relatively straight forward using shale volume, porosity, water saturation, and permeability calculated from a petrophysical log analysis. Calibration to core data will help make the results believable.
 

Carbonate reservoirs are often more heterogeneous laterally than sandstone reservoirs. Porosity-permeability modeling within carbonates is possible provided it is carried out at an appropriate scale. Results should not be extended beyond individual flow units. Petrographic data, coupled with core analysis data, are used to generate log signature transforms that help to partition matrix (ineffective) porosity from effective (useful) porosity.